Trends in Alcohol Use Before and During the COVID-19 Pandemic Among Women Living With and Without HIV in the United States (2017–2022)

Hannah Tierney, Yifei Ma, Fan Xia, Aruna Chandran, Mirjam-Colette Kempf, Lauren F. Collins, Jack DeHovitz, Ralph J. DiClemente, Michelle Floris-Moore, Audrey L. French, Deborah L. Jones, Anjali Sharma, Amanda B. Spence, Judith A. Hahn, Jennifer C. Price, Phyllis C. TienAIDS 2025 · DOI

Abstract

Data on alcohol use trends in women living with HIV (WLWH) are lacking. We examine data collected before and across the COVID-19 pandemic (2014-2022) in WLWH and socio-demographically similar women living without HIV (WLWOH) (n = 1564; 67% WLWH) who self-reported alcohol use to establish and compare trends. Women were categorized into baseline groups based on use pattern from 10/2014 to 09/2017: abstinence (no alcohol at any visit), non-heavy drinking (0 heavy drinking visits), some heavy drinking (1-4 visits), and persistent heavy drinking (5-6 visits) with heavy drinking defined as > 7 drinks/week or > 3 drinks/day. A linear mixed model (LMM) was used to assess for changes in average number of drinks/week (ADW) by time segment: pre-pandemic (10/2017-09/2019), early pandemic (03/2020-05/2021), and late pandemic (06/2021-09/2022). Time, baseline group, and HIV status were included as interactions. In the LMM, WLWH with persistent heavy drinking histories had lower ADW compared to WLWOH. There were no significant differences in ADW in the other groups by HIV status and no significant interactions between HIV status and the time segments. The largest change in ADW occurred among the persistent heavy drinking group, with decreases at pandemic onset followed by increases into 2021. Overall, WLWH reported less alcohol use than WLWOH, but changes in alcohol use trends from 10/2017 to 09/2022 were similar regardless of HIV status. This suggests that the pandemic did not uniquely impact alcohol use among WLWH, rather socio-economically disadvantaged women generally. These findings support the need for greater research in the role of socioeconomic disruption on alcohol use among women.

Introduction

Unhealthy alcohol consumption is a leading cause of preventable mortality, with rates rising steadily among women in the United States. Historically, women living with HIV (WLWH) have been identified as a population reporting higher rates of alcohol use compared to the general public. While early pandemic-era studies suggested a general spike in drinking among adults, many of these findings were based on short-term, cross-sectional data that failed to account for long-term historical drinking trajectories.

This study addresses this gap by examining a longitudinal timeline from 2017 to 2022. The primary objective was to determine if the COVID-19 pandemic uniquely impacted the alcohol consumption of WLWH compared to socio-demographically similar women living without HIV (WLWOH), or if behavioral shifts were primarily dictated by pre-existing drinking habits.

Methods

The study utilized data from the Women’s Interagency HIV Study (WIHS), now part of the MACS/WIHS Combined Cohort Study (MWCCS), which represents the largest and longest-running HIV cohort of women in the U.S. The analysis included 1,564 participants (67% WLWH).

Study Design and Stratification

To provide a nuanced view of behavioral changes, researchers first established a baseline drinking profile for each participant using data from October 2014 to September 2017. Participants were stratified into four groups:

  • Abstinence: No alcohol reported at any visit.
  • Non-heavy drinking: Reported alcohol use but zero instances of heavy drinking.
  • Some heavy drinking: Heavy drinking reported at 1–4 visits.
  • Persistent heavy drinking: Heavy drinking reported at 5–6 visits.

Heavy drinking was defined as consuming more than 7 drinks per week or more than 3 drinks in a single day.

Timeline of the study periods, including baseline categorization and pandemic-related follow-up segments.

Statistical Analysis

Researchers employed Linear Mixed Models (LMM) with piecewise regression to analyze trends in average drinks per week (ADW) across three distinct time segments: pre-pandemic (10/2017–09/2019), early pandemic (03/2020–05/2021), and late pandemic (06/2021–09/2022). This approach allowed the team to account for within-subject correlation and prevent stable habits in the majority from masking volatile shifts in high-risk subgroups.

Key Results and Insights

The study revealed that the pandemic's impact on drinking was dictated almost entirely by historical behavior rather than HIV status.

The "Heavy Drinker" Rollercoaster

The most dramatic fluctuations were observed in the persistent heavy drinking group. Counterintuitively, these individuals did not simply drink more when the pandemic began. Instead, they reported a decrease in consumption at the onset—likely due to the closure of bars, social isolation, and financial strain. However, as restrictions lifted in 2021, their consumption spiked significantly before trending downward again in 2022.

HIV Status and Consumption Volume

WLWH actually reported less alcohol use overall compared to the WLWOH control group. Among persistent heavy drinkers, WLWH consumed an average of 8.75 fewer drinks per week than WLWOH (p < 0.001). The researchers hypothesize that this may be due to higher levels of primary care engagement and health counseling among WLWH within this long-standing cohort.

Shared Pandemic Trajectories

Crucially, there were no significant interactions between HIV status and the pandemic time segments. Both WLWH and WLWOH followed remarkably similar trajectory shapes across the timeline. This suggests that the pandemic did not exert a unique pressure on WLWH; rather, the observed shifts were characteristic of socio-economically disadvantaged women generally.

Discussion and Limitations

These findings highlight the danger of relying on aggregate data, which would have masked the sharp spikes and drops seen in high-risk subgroups. The study underscores that historical drinking patterns are the most reliable predictor of how individuals respond to large-scale social disruptions.

Limitations

The study is limited by its reliance on self-reported data, which is susceptible to recall and social desirability biases. Furthermore, the transition from in-person to phone interviews during the pandemic may have influenced reporting behaviors. Finally, the study did not directly measure specific pandemic-era stressors, such as job loss or acute anxiety, which could provide further context for the observed fluctuations.

Future Work

Future research should explore the specific protective factors—such as resilience and healthcare engagement—that kept consumption lower among WLWH and examine whether the pandemic has left a permanent mark on alcohol use trajectories over the coming decade.

Summary updated

Abstract

Data on alcohol use trends in women living with HIV (WLWH) are lacking. We examine data collected before and across the COVID-19 pandemic (2014-2022) in WLWH and socio-demographically similar women living without HIV (WLWOH)(n = 1564; 67% WLWH) who self-reported alcohol use to establish and compare trends. Women were categorized into baseline groups based on use pattern from 10/2014 to 09/2017: abstinence (no alcohol at any visit), non-heavy drinking (0 heavy drinking visits), some heavy drinking (1-4 visits), and persistent heavy drinking (5-6 visits) with heavy drinking defined as > 7 drinks/week or > 3 drinks/day. A linear mixed model (LMM) was used to assess for changes in average number of drinks/week (ADW) by time segment: pre-pandemic (10/2017-09/2019), early pandemic (08/2020-05/2021), and late pandemic (06/2021-09/2022). Time, baseline group, and HIV status were included as interactions. In the LMM, WLWH with persistent heavy drinking histories had lower ADW compared to WLWOH. There were no significant differences in ADW in the other groups by HIV status and no significant interactions between HIV status and the time segments. The largest change in ADW occurred among the persistent heavy drinking group, with decreases at pandemic onset followed by increases into 2021. Overall, WLWH reported less alcohol use than WLWOH, but changes in alcohol use trends from 10/2017 to 09/2022 were similar regardless of HIV status. This suggests that the pandemic did not uniquely impact alcohol use among WLWH, rather socio-economically disadvantaged women generally. These findings support the need for greater research in the role of socioeconomic disruption on alcohol use among women.
BibTeX
@inproceedings{tierney2025trends,
  author = {Hannah Tierney and Yifei Ma and Fan Xia and Aruna Chandran and Mirjam-Colette Kempf and Lauren F. Collins and Jack DeHovitz and Ralph J. DiClemente and Michelle Floris-Moore and Audrey L. French and Deborah L. Jones and Anjali Sharma and Amanda B. Spence and Judith A. Hahn and Jennifer C. Price and Phyllis C. Tien},
  title = {{Trends in Alcohol Use Before and During the COVID-19 Pandemic Among Women Living With and Without HIV in the United States (2017–2022)}},
  booktitle = {AIDS},
  year = {2025},
  doi = {10.1007/s10461-025-04875-9},
}
Figure